Neuropixels Analysis API 参考

# API 参考 按模块组织的 neuropixels_analysis 函数快速参考。 ## 核心模块 (Core Module) ### load_recording python npa.load_recording( path: str, format: str = 'auto', # 'spikeglx', 'openephys', 'nwb' stream_id: str = None, # e.g., 'imec0.ap' ) -> Recording 从各种格式加载 Neuropixels 记录。 ### run_pipeline python npa.run_pipeline( recording: Recording, output_dir: str, sorter: str = 'kilosort4', preprocess: bool = True, correct_motion: bool = True, postprocess: bool = True, curate: bool = True, curation_method: str = 'allen', ) -> dict 运行完整的分析工作流。返回包含所有结果的字典。 ## 预处理模块 (Preprocessing Module) ### preprocess python npa.preprocess( recording: Recording, freq_min: float = 300, freq_max: float = 6000, phase_shift: bool = True, common_ref: bool = True, bad_channel_detection: bool = True, ) -> Recording 应用标准预处理链。 ### detect_bad_channels python npa.detect_bad_channels( recording: Recording, method: str = 'coherence+psd', **kwargs, ) -> list 检测并返回坏通道 ID 列表。 ### apply_filters python npa.apply_filters( recording: Recording, freq_min: float = 300, freq_max: float = 6000, filter_type: str = 'bandpass', ) -> Recording 应用频率滤波器。 ### common_reference python npa.common_reference( recording: Recording, operator: str = 'median', reference: str = 'global', ) -> Recording 应用公共参考 (CMR/CAR)。 ## 运动模块 (Motion Module) ### check_drift python npa.check_drift( recording: Recording, plot: bool = True, output: str = None, ) -> dict 检查记录是否存在漂移。返回漂移统计信息。 ### estimate_motion python npa.estimate_motion( recording: Recording, preset: str = 'kilosort_like', **kwargs, ) -> dict 估算运动而不应用校正。 ### correct_motion python npa.correct_motion( recording: Recording, preset: str = 'nonrigid_accurate', folder: str = None, **kwargs, ) -> Recording 应用运动校正。 **预设 (Presets):** - `'kilosort_like'`: 快速、刚性校正 - `'nonrigid_accurate'`: 较慢，适用于严重漂移 - `'nonrigid_fast_and_accurate'`: 平衡选项 ## 排序模块 (Sorting Module) ### run_sorting python npa.run_sorting( recording: Recording, sorter: str = 'kilosort4', output